A major challenge facing cancer treatment is tumor heterogeneity and selection of drug resistant clones. It is widely accepted that tumors possess large amounts of genetic heterogeneity due to branching subclonal evolution. Recently, it was discovered that heterogeneity also exists within isogenic tumor populations. Cellular heterogeneity has been attributed to stochastic processes intrinsic to gene expression and to epigenetic variability. Phenotypic switching has been observed in breast cancer cells that can reestablish equilibrium of subpopulations of stem, luminal, and basal phenotypes following perturbation. Moreover, many NSCLC patients show a reversible drug tolerance to Tyrosine Kinase Inhibitors. I hypothesize that distinct, semi- heritable epigenetic states underlie cellular heterogeneity in tumors. I also predict this epigenetic plasticity exists due to incomplete resetting of histone modifications. Cellular heterogeneity has been a relatively unexplored phenomena due to a lack of technologies that can trace the fates of individual tumor cells. In this proposal I examine the mechanisms of cellular heterogeneity in tumors using a novel lineage tracking system, TRACER, where each cell writes its own unique barcode while recording the evolutionary history of ancestral cells. I propose to combine this transformative technology with pooled lentiviral CRISPR/gRNA libraries to provide a single cell resolution previously unavailable within pooled gene perturbation studies. Using these tools, I will quantitate the interconversion between semi-stable epigenetic states and define the heritability of these states while under selective pressures. The proposed experiments will illuminate the mechanisms of cellular heterogeneity and reversible drug tolerance in cancer cells and identify epigenetic drug resistance genes. The long term goal of TRACER studies is to identify the mechanisms that control the evolutionary trajectories of tumors and inform the design of drugs aimed at slowing or directing predictable paths of tumor evolution, with the ultimate goal of converting cancer from a lethal to a chronic disease.